INDEX

Explanations

the word "changes" at different levels of magnitude

oai_token-act-pair · gpt-3.5-turbo

references to modifications or alterations in policies or regulations

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.5.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ographies

-0.72

AFB

-0.70

Äĩ

-0.69

 DRAGON

-0.67

ZE

-0.65

ãĥ¥

-0.65

amina

-0.65

Write

-0.65

ILLE

-0.65

zees

-0.61

POSITIVE LOGITS

uits

0.99

 wrought

0.98

 effected

0.89

hops

0.87

ettings

0.83

undown

0.81

hift

0.79

oodoo

0.78

hooting

0.76

oldown

0.75

Activations Density 0.039%

the word "changes" at different levels of magnitude

references to modifications or alterations in policies or regulations

No Comments

No Known Activations